The RICS AI Standard For Chartered Surveyors, Explained

In March 2026, the Royal Institute of Chartered Surveyors published new professional standards - Responsible Use of Artificial Intelligence in Surveying Practice.

With a lack of government-led stand-alone legislation for AI in the UK, many professional bodies and societies have opted to introduce their own sector-specific standards, providing a code for members that encourages and educates on responsible use of AI. RICS issued their AI Standard earlier this year, and it’s essential that all RICS members are not only aware of this but really understand what the guidance is.

In this guide - The RICS AI Standard For Chartered Surveyors - we’ll break down the key points, unravel the advice and explain what you, and/or your company, needs to be doing to comply with this standard.

Why Is The RICS AI Standard Important?

It’s important to clarify that issuing a standard regarding AI does not mean the organisation is seeking to discourage the use of AI tools or advise against harnessing new technologies to elevate the standard of work and increase productivity. Rather, it is there to help surveyors mitigate risk, stay compliant with other regulations and maintain professional standards as we all adapt to these emerging technologies.

Almost overnight, it seemed, the future arrived. Generative AI followed closely by agentic AI, capable of high-speed ‘thinking’, content generation and even decision-making. If it seems impossible to keep up with the capabilities of AI, that’s because it is. But whilst we’re busy trying to, we mustn’t allow this pursuit to impact on our professional responsibilities or lead us to neglect best practices.

The RICS AI Standard is important because it provides a roadmap for surveyors, one that’s fully onboard with incorporating AI into the way we work but flags up potential obstacles and gives direction on safe and ethical use.

Moreover, if you’re a Chartered Surveyor, then you’re obliged to follow these new standards. So, let’s go through them -

Human Oversight, Accountability & Professional Judgment

The RICS strongly emphasises the importance of human oversight. In both this guidance and in other regulations around compliance, the buck stops with the operator and not the AI tool, so you are accountable.

The standard is clear that professional judgement, made up of knowledge, skills, experience and professional scepticism, has to be applied to every AI output that materially affects a service. The reliability of the output must be decided by a qualified surveyor and must be documented in writing.

That written decision needs to cover any assumptions made, the key concerns about reliability, why those concerns exist, and whether anything could reduce them. This should lead to a conclusion on whether the output can reasonably be relied on for its intended purpose. Where the answer is no, the firm must disclose that to the client in writing, along with the reasoning behind it.

Of course, it’s not always possible or necessary to scrutinise an AI tool’s every individual output when utilising automated or high-volume AI, but this does not reduce your culpability for errors; therefore, the RICS instructs firms to carry out regular but randomised dip samples to ensure quality across the batch.

Evidence of human review is crucial here - the organisation must be able to prove that human review actually happened, and happened properly. This could safeguard you somewhat if an AI output is later challenged.

Understanding and Implementing the RICS Mandatory Requirements

The standard is built around five sections: baseline knowledge, practice management, procurement, output reliance and assurance, and development.

Within these, the RICS distinguishes between what is mandatory and what is recommended as follows –

• Where the word "must" is used, requirements are mandatory

• Where the word "should" is used, these are recommendations, and alternative approaches may be implemented so long as they achieve the same or a better outcome

In regulatory or disciplinary proceedings, and potentially in legal proceedings, these standards will be taken into account when deciding whether a member or firm acted responsibly.

The core duties break down as follows –

• Members using AI to deliver surveying services must maintain a basic working understanding of how different AI systems function, their limitations and failure modes, the risk of erroneous output, the risk of bias, and the data risks involved.

• Firms must put practice management structures in place, including data governance, system governance and risk management.

• Anyone relying on an AI output for a service with material impact must apply professional judgement and document their reasoning.

• Transparency must be practised so clients are told, upfront and in writing, when and how AI is being used.

  
  

Key to the standard is recognising that AI should not be a tool elevated above professional judgement - it’s the surveyor's judgement, not the system's output, that remains crucial.

Updating Terms & Conditions and Client Engagement Letters

Transparency sits at the heart of this standard, and RICS makes it explicit that if AI has a material impact on how you deliver a surveying service, your client needs to know before you start work, not after. This isn't a footnote requirement. It must be built into your terms of engagement, contractual documents and service agreements, in writing.

Specifically, these documents now need to set out when AI will be involved in delivering the service, which parts of that process AI is incorporated into, the extent of your professional indemnity cover for AI use, how a client can raise concern or complaints regarding your use of AI and if they feel an AI system has negatively affected them. They should also be informed about how they might opt out of AI use altogether, if that is an option.

This standard will mean a genuine rewrite of your standard documentation, not a bolt-on clause. Every point above needs to be addressed explicitly, and every client needs to see it before any work begins.

For help incorporating AI disclosure and directives into your client terms and conditions, please reach out to Aubergine Legal who offer this service.

AI Governance, Risk Registers & Documentation

Before using any AI system that will materially affect service delivery, RICS-regulated firms must carry out and record a written assessment of whether AI is actually the right tool for the job.

This assessment should weigh up –

• The surveying services provided

• The nature of the task

• Available alternatives

• Environmental and stakeholder impact

• Data risks

• The likelihood of inaccurate or biased output

• The consequences of inaccurate or biased output

This can sit within a standalone policy or a periodically reviewed statement, but must exist in writing.

Firms must also maintain a written AI system register, logging every AI system in material use, its purpose, when it was first used, and when its suitability will next be reviewed.

Then there's the risk register itself, which must document overarching risks including inherent bias, erroneous outputs, gaps in the underlying training data, and how the firm retains or uses data it feeds into the system. For each risk, firms need to record a description, likelihood and impact, mitigation plan, risk appetite, and a RAG rating or similar. This register isn't a one-off exercise. It must be reviewed and updated at least quarterly by whoever is responsible for the firm's AI decisions.

Data Protection & Confidentiality Controls

The RICS standard encourages members to treat AI-related data risk as a distinct category, separate from the data risks surveyors are already used to managing.

Already, members must store private and confidential data securely, restrict access to staff who genuinely need it, and train those staff at least annually on privacy and confidentiality risks, adhering to UK GDPR and other professional guidelines. However, the unique risks posed by AI must also be approached in the same manner, but as a separate factor.

AI introduces new risks, and further mitigation is often needed. In relation to AI, data should be prepared and imputed for use in a way that protects privacy, such as through anonymisation, wherever practical.

The default position on uploading confidential information into an AI system is not to do so. Where it is deemed necessary, two conditions must be met -

• The firm has written consent from affected stakeholders in advance

• The firm has taken reasonable steps to ensure the upload doesn't pose an unacceptable risk

This sits alongside, not instead of, existing obligations under UK GDPR, your confidentiality duties to clients, and whatever your professional indemnity insurer requires from you. None of those obligations is lessened because of AI use. If anything, this standard demands more controls and security.

AI Procurement, Vendor Due Diligence & Contracting

Before procuring any third-party AI system that will affect service delivery, firms must carry out detailed due diligence. That means actively requesting information from the supplier in writing, following up in writing where needed, recording what's provided, and testing the system for fitness for purpose.

Usually, we’d advise thoroughly reading and understanding the tool’s service agreement, but the RICS isn't satisfied with this alone. Their standard requests an active, initiated and evidenced exchange with the supplier, rather than relying on the terms the supplier made available.

At a minimum, the written request must cover the system's environmental impact, who was involved in developing it, its compliance with data and confidentiality law, what permissions were obtained for any personal data used in training, the accuracy, relevance and diversity of the training data, including known gaps and bias risks, and the extent of the supplier's liability.

If a supplier gives limited or no answer to any of this, that doesn’t necessarily mean the AI cannot be used, but you must then identify the risks created by the gap and log them in its risk register, rather than proceeding on the strength of the contract alone.

In practice, this means procurement can't be a quick sign-off. Instead, a documented interrogation of the vendor is required before any AI system goes near client work.

In Summary

The RICS standard doesn’t seek to slow down AI adoption within the profession. It's about proving that adoption hasn't come at the cost of the professional judgment clients rely on.

Conveyance Surveyors must also be -

●       Keeping risk registers and reviewing these regularly

●       Documenting decisions on output reliability

●       Conducting proper due diligence on vendors

●       Updating client documents to disclose and detail AI use

Responsible Use of Artificial Intelligence in Surveying Practice is a standard that puts processes in place that keep us from assuming a system is safe simply because it's convenient. If you're a Chartered Surveyor, this isn't recommended reading - it's the standard you're already required to meet.

Download our AI Compliance Checklist for Businesses to help you assess where you are and what else you might need to do. Aubergine Legal also offer a comprehensive AI Legal Toolkit, which navigates you through helpful checklists and provides necessary template policies and inserts for your legal documents, to ensure compliance with data protection and IP laws.  Finally bepoke audits of tools and help writing policies, logs and updates to your client terms and privacy policies is a service that we offer – so get in touch if you need help navigating all this – it’s what we do for clients.